Biochemistry 2009, 48, 3801–3803
3801
DOI:10.1021/bi900308c
Monitoring the Interaction of a Single G-Protein Key Binding Site with Rhodopsin
Disk Membranes upon Light Activation†
::
Tai-Yang Kim,‡ Hiroshi Uji-i,§ Martina Moller,‡ Benoit Muls,§ Johan Hofkens,§
and Ulrike Alexiev*,‡
‡
::
Physics Department, Freie Universitat Berlin, Arnimallee 14, D-14195 Berlin, Germany, and §Division of Molecular and
Nano Materials, Katholieke Universiteit Leuven, Celestinienlaan 200F, 3001 Leuven, Belgium
Received February 22, 2009. Revised Manuscript Received March 19, 2009
ABSTRACT: Heterotrimeric G-proteins interact with
their G-protein-coupled receptors (GPCRs) via key
binding elements comprising the receptor-specific
C-terminal segment of the R-subunit and the lipid
anchors at the R-subunit N-terminus and the γ-subunit
C-terminus. Direct information about diffusion and
interaction of GPCRs and their G-proteins is mandatory for an understanding of the signal transduction
mechanism. By using single-particle tracking, we show
that the encounters of the R-subunit C-terminus with
the GPCR rhodopsin change after receptor activation.
Slow as well as less restricted diffusion compared to the
inactive state within domains 60-280 nm in length was
found for the receptor-bound C-terminus, indicating
short-range order in rhodopsin packing.
G-Protein-coupled receptors (GPCRs) constitute a large
superfamily of membrane receptors responsible for the
transduction of various external stimuli (1). In rhodopsin,
a prototype of GPCR subfamily A, the reception of light
and subsequent signal mediation across the membrane
leads to conformational changes in the receptor, facilitating
molecular recognition between the active MetarhodopsinII (Meta-II) intermediate and the G-protein transducin on
the cytosolic face of the receptor (2). The specialization of
rhodopsin in phototransduction is reflected in its structure
and cellular arrangement: High concentrations of the
receptor in the stacked disks located in the rod outer
segments (ROS) are combined with a lack of most of the
other proteins involved in cellular function and ultrafast
switching from the covalently bound inactive (11-cis-retinal) to the active ligand (all-trans-retinal). The packing
properties of rhodopsin in the disk membrane and the
diffusional characteristics of transducin will affect the
timing of events in the complex signal transduction cascade.
Simulations of stochastic encounters between light-activated rhodopsin and transducin reveal a clear dependence
on rhodopsin packing and lateral diffusion of transducin; a
respective high lateral diffusion coefficient of transducin
†
This work was partially supported by the German Science
Foundation, the FWO, Katholieke Universiteit Leuven, and the
Flemish government.
*To whom correspondence should be addressed. Phone: ++4930-838-55157. Fax: ++49-30-838-56510. E-mail: alexiev@physik.
fu-berlin.de.
© 2009 American Chemical Society
seems to be sufficient to obtain the fast responses required,
even at high concentrations of immobile rhodopsins (3).
However, the packing of rhodopsin in the membrane and
the related diffusion coefficients are controversially discussed. While early experiments indicated that rhodopsin
diffuses freely in the lipid bilayer of the disk membrane (4),
recent atomic force microscopy data revealed rows of
rhodopsin dimers in a paracrystalline lattice (5), an arrangement which seems to conflict with a freely diffusing
rhodopsin. Thus, direct information about the diffusion
and interaction properties of rhodopsin and transducin is
mandatory for deducing the underlying microscopic principle of action. To the best of our knowledge, direct
measurements of lateral diffusion behavior of transducin
on the single-molecule level were not reported, so we set out
to implement a setup based on single-molecule tracking to
follow the encounters and lateral diffusion of transducin
and its binding elements with the disk membrane.
Here we performed a proof-of-principle experiment to
show the feasibility of the single-molecule spectroscopy
approach to studying the time dependence of the lighttriggered interaction between rhodopsin and transducin.
For a clear-cut discrimination between the different interaction modes in the dark and light-activated state of
rhodopsin, we took advantage of the fact that transducin
(GT)-derived peptides of the GTR-subunit C-terminus
recognize the active form of the receptor and can replace
holotransducin in stabilizing Meta-II (6). In particular, the
peptide derived from the extreme C-terminus of the GTRsubunit has supposedly no affinity for dark rhodopsin (7)
but high affinity for light-activated rhodopsin (6,7). Thus,
this peptide is expected to show membrane interaction only
after light activation, in contrast to holotransducin with its
lipid anchors. A recent crystal structure highlights the
interaction of the transducin peptide with an active state
of the receptor (8) (Figure 1C).
For our experiments, we used a fluorescently labeled
variant of the C-terminal segment of the GTR-subunit,
pGTR-F (F is the fluorescence label Atto647N). In the
initial experiment, we followed the interaction of pGTR-F
with inactive (dark) rhodopsin membranes using singlemolecule wide-field microscopy (9). As rhodopsin membrane sheets are known to form well-defined layer
Published on Web 3/20/2009
pubs.acs.org/Biochemistry
3802
Biochemistry, Vol. 48, No. 18, 2009
FIGURE 1: Fluorescence images of single pGTR-F. (A) Interaction with dark
rhodopsin (ROS) membranes. (B) Image series from the time-resolved interaction of pGTR-F with light-activated disk membranes. Same area as in panel
A but with a reduced amount of fluorescent pGTR-F. (C) Structure of opsin
(cyan) in complex with the C-terminal peptide of the transducin R-subunit
(red) (8), together with receptor-unbound fluorescently labeled peptides (blue).
(D) Kinetics of Meta-II formation (20) and increase in the number (N) of
interacting pGTR-F molecules after rhodopsin light activation. Conditions:
130 mM NaCl, 10 mM MES, pH 6, 19 C.
structures (10), we deposited disk membrane layers at the
surface of a glass coverslip in a buffer-filled sample compartment. The fluorescently labeled pGTR molecules are
visible as bright spots when added to the nonlabeled
rhodopsin disk membranes. With increased pGTR-F concentrations, areas of high pGTR-F localization were
observed (Figure 1A), consistent with the outline of individual membrane patches. In control experiments without membranes, we observed a homogeneous interaction
pattern (data not shown). To allow for a visualization of
the expected increase in the number of pGTR-F molecules
interacting with the activated rhodopsin, we reduced the
number of fluorescent pGTR-F molecules sufficiently
(Figure 1B, first image). Figure 1B presents an image series
from the first time-resolved observation of the interaction
and diffusion characteristics of pGTR-F with disk membranes and their changes upon light activation. An integration time of 31 ms per frame (32 Hz) was used. The
image series starts with the frame before the green flash
activates rhodopsin and clearly shows that rhodopsin
activation triggers an increase in the number of interacting
pGTR-F molecules. This increase was absent without
rhodopsin membranes (Figure 1D). Within the first frame
after the flash, the highest number of pGTR-F molecules
was observed at the membrane surface. Thus, initial binding occurs within the formation time of Meta-II
(Figure 1D). This is fast compared to the binding kinetics
in the time range of seconds observed for mass-tagged Cterminal segments in classical light scattering experiments (7), but comparable with the binding time constant
of GT (40 ms) in earlier reports (11). The observation of
light-induced binding of single pGTR-F molecules to rhodopsin disk membranes was a crucial test for the successful
application of single-molecule fluorescence spectroscopy.
To further quantify the wide-field single-molecule fluorescence experiments, we analyzed the mobility of pGTR-F
molecules. As the GT-derived peptide interacts tightly with
the receptor after rhodopsin light activation (6), it is
expected to report on the mobility of the active Meta-II
molecule. Mean square displacements (Ær2æ) obtained from
pGTR-F trajectories are presented for several molecules
before and after rhodopsin light activation (Figure 2B,E).
The distribution of the distances r (step length) a molecule moves in a given time allows for the detection of
subpopulations in nonhomogeneous populations (12, 13).
Nonhomogeneous populations are suspected from the Ær2æ
plot and are indicated in Figure 2B,E with different
colors. The step length distributions for inactive and
light-activated rhodopsin membranes can be satisfactorily
described by a biexponential fit, yielding two subpopulations (Figure 2C,F). In the dark state of rhodopsin, we
observed that ∼70% of the pGTR-F molecules interact
only for one frame with the disk membranes (Figure 2A),
consistent with the vanishing affinity of the peptidic binding element for inactive rhodopsin. Less than 20% of the
molecules contribute to the trajectories and the step length
distribution shown in Figure 2B,C. The time dependence
of the coefficients from the distribution functions reveals
that the r2 values of the major fraction are constant and
close to the displacement accuracy of our experiment, thus
indicating immobile pGTR-F. As a fit to the decaying
residence time histogram (Figure S3A) results in a time
constant of 49 ms, those molecules which interact for
two frames were analyzed separately. They display a
higher mobility: r21 = (6.0 ( 1.2) 10-3 μm2, and r22 =
(47 ( 1.2) 10-3 μm2. According to the relationship r2 =
4Dt for normal diffusion, this yields the following diffusion
constants: D1 = 0.05 μm2/s, and D2 = 0.38 μm2/s. The
situation is different after light activation of rhodopsin.
The percentage of molecules, which interact for only one
frame, is reduced to ∼40% (Figure 2D), consistent with the
fact that the majority of pGTR-F tightly binds to active
rhodopsin. However, rhodopsin light activation surprisingly results in an increased heterogeneity in pGTR-F
mobility (Figure 2E). More mobile components are evident from the step length distribution histogram
(Figure 2F, orange fit) and the corresponding cumulative
distribution function (Figure 3A). The fractional amplitude of the fast component is increased from ∼0.2 in the
dark to ∼0.5 after light activation (Figure 3A,B). While for
FIGURE 2: Residence times and time dependence of Ær2æ. Dark rhodopsin
membranes. (A) Residence times of interacting pGTR-F with dark disk
membranes. (B) Mean square displacements: thick lines, mean of respectively
colored time traces; inset, selected tracks. (C) Step length distribution. Fit
(dashed line): r21 = 0.0009 ( 0.0001 μm2 (green); r22 = 0.003 ( 0.001 μm2 (blue).
Light-activated rhodopsin membranes. (D) Residence times of interacting
pGTR-F with light-activated disk membranes. (E) Mean square displacements:
bold, mean of respectively colored time traces; inset, selected tracks. (F) Step
length distribution. Fit: r21 = 0.001 ( 0.0001 μm2 (green); r22 = 0.005 ( 0.001
μm2 (orange).
Rapid Report
FIGURE 3: Time dependence of distribution function coefficients. (A) Prob-
ability distribution function for square displacements of pGTR-F interacting
with dark (black; fit, r21 = 0.0009 μm2, r22 = 0.0037 μm2, Α1 = 0.81) and lightactivated (red; fit, r21 = 0.001 μm2, r22 = 0.0047 μm2, Α1 = 0.42) rhodopsin
membranes. For immobile pGTR-F, r2 = 0.0002 μm2. (B) Time lag dependence of amplitude A1 (slow fraction) after rhodopsin light activation. (C)
Time lag dependence of r2 (slow component) after rhodopsin light activation
from a selected experiment (fit, L1 = 61 ( 2 nm, D0 = 0.01 μm2/s). (D) Time
lag dependence of r2 (fast component) after rhodopsin light activation from a
selected experiment (fit, L2 = 130 ( 2 nm, D0 = 0.1 μm2/s). The dotted
horizontal line in panels C and D represents the mean displacement accuracy.
Biochemistry, Vol. 48, No. 18, 2009
3803
Taken together, our single-molecule tracking data support the conclusion that membrane microdomains play a
role in the organization of GPCRs. In the case of rhodopsin, single light-activated receptors diffuse in membrane
domains of different sizes at a time scale of milliseconds to
seconds. This suggests disk membrane inhomogeneity and
is compatible with a semiordered packing of rhodopsin
over a time range which corresponds to the millisecond
activation time scale of rhodopsin. The successful visualization of the light-induced interaction of a single transducin binding element with rhodopsin opens the possibility to
further investigate diffusion characteristics and interactions of holotransducin and its subunits on the singlemolecule level to provide deeper insights into the regulation of receptor-G-protein activation. Within a greater
scheme, the interaction and the mutual interference with
molecules controlling the multistep shutoff of activated
rhodopsin can be directly visualized, and their dynamics in
the signaling cascade can be studied in the future.
ACKNOWLEDGMENT
We thank Dr. Sebastian Haase for performing the test
with simulated data.
SUPPORTING INFORMATION AVAILABLE
the dark rhodopsin membranes strongly confined diffusion and immobilization of interacting pGTR-F were observed, the time dependence of the r2 values after light
activation indicates less restricted diffusion (Figure 3C,D).
Restricted diffusion behavior (14) may be explained by a
confined diffusion model (15). The confinement lengths
(L) are between 60 nm (subpopulation 1) and 280 nm
(subpopulation 2), depending on the sample, with the
following initial diffusion constants: D0 = 0.01 μm2/s,
and D0 = 0.1 μm2/s, respectively. More interestingly,
subpopulation 1 displays an increase in r2 with time,
starting about 200 ms after light activation of rhodopsin.
This suggests a change to normal diffusion with a D of
0.002 μm2/s (Figure 3C). This phenomenon was observed
in all experiments (three different samples). We propose
that the observed transition to normal diffusion after
rhodopsin light activation is caused by slow diffusion of
the receptor detected through rhodopsin-bound pGTR-F
molecules. The change in diffusion behavior is probably
induced by alterations in rhodopsin packing, supported by
the observed increase in confinement length after light
activation, and by structural changes in activated rhodopsin. Studies on rhodopsin oligomerization suggest that
rhodopsin tends to oligomerize, and simulations predict
short-range order in rhodopsin membrane assembly (16,
17). A recent fluorescence recovery after photobleaching
study (18) on GT and rhodopsin in tadpole rods found high
mobility for activated, G-protein-devoid rhodopsin, but
sequestration into cholesterol-dependent microdomains
for the GT-bound receptor. The diffusion constant of the
latter (0.05 μm2/s) is within the range of the D0 values
found here for the activated receptor-bound pGTR-F. Only
those pGTR-F which briefly interact with the inactive
membrane display higher mobility (D ∼ 0.4 μm2/s). We
suggest that these high-mobility molecules probably interact with the lipid phase, as mobile G-protein lipid anchors
displayed diffusion coefficients in the same range (19).
Details of experimental procedures and data analysis.
This material is available free of charge via the Internet at
http://pubs.acs.org.
REFERENCES
1. Bockaert, J., and Pin, J. P. (1999) EMBO J. 18, 1723–1729.
2. Sakmar, T. P., Menon, S. T., Marin, E. P., and Awad, E. S. (2002)
Annu. Rev. Biophys. Biomol. Struct. 31, 443–484.
3. Dell’Orco, D., and Schmidt, H. (2008) J. Phys. Chem. B 112, 4419–
4426.
4. Poo, M., and Cone, R. A. (1974) Nature 247, 438–441.
5. Fotiadis, D., Liang, Y., Filipek, S., Saperstein, D. A., Engel, A., and
Palczewski, K. (2003) Nature 421, 127–128.
6. Kisselev, O. G., Meyer, C. K., Heck, M., Ernst, O. P., and Hofmann,
K. P. (1999) Proc. Natl. Acad. Sci. U.S.A. 96, 4898–4903.
7. Herrmann, R., Heck, M., Henklein, P., Henklein, P., Kleuss, C.,
Hofmann, K. P., and Ernst, O. P. (2004) J. Biol. Chem. 279, 24283–
24290.
8. Scheerer, P., Park, J. H., Hildebrand, P. W., Kim, Y. J., Krauss, N.,
Choe, H. W., Hofmann, K. P., and Ernst, O. P. (2008) Nature 455,
497–502.
::
9. Rocha, S., Hutchison, J. A., Peneva, K., Herrmann, A., Mullen, K.,
Skjøt, M., Jørgensen, C. I., Svendsen, A., De Schryver, F. C.,
Hofkens, J., and Uji-I, H. (2009) ChemPhysChem 10, 151–161.
10. Fitter, J., Ernst, O. P., Hauss, T., Lechner, R. E., Hofmann, K. P.,
and Dencher, N. A. (1998) Eur. Biophys. J. 27, 638–645.
::
11. Bennett, N., Michel-Villaz, M., and Kuhn, H. (1982) Eur. J.
Biochem. 127, 97–103.
::
12. Kues, T., Dickmanns, A., Luhrmann, R., Peters, R., and Kubitscheck,
U. (2001) Proc. Natl. Acad. Sci. U.S.A. 98, 12021–12026.
::
13. Schutz, G. J., Schindler, H., and Schmidt, T. (1997) Biophys. J. 73,
1073–1080.
14. Saxton, M. J., and Jacobson, K. (1997) Annu. Rev. Biophys. Biomol.
Struct. 26, 373–399.
15. Kusumi, A., Sako, Y., and Yamamoto, M. (1993) Biophys. J. 65,
2021–2040.
16. Periole, X., Huber, T., Marrink, S.-J., and Sakmar, T. P. (2007) J.
Am. Chem. Soc. 129, 10126–10132.
17. Botelho, A. V., Huber, T., Sakmar, T. P., and Brown, M. F. (2006)
Biophys. J. 91, 4464–4477.
18. Wang, Q., Zhang, X., Zhang, L., He, F., Zhang, G., Jamrich, M.,
and Wensel, T. G. (2008) J. Biol. Chem. 283, 30015–30024.
19. Perez, J.-P., Segura, J.-M., Abankwa, D., Piguet, J., Martinez, K. L.,
and Vogel, H. (2006) J. Mol. Biol. 363, 918–930.
::
20. Alexiev, U., Rimke, I., and Pohlmann, T. (2003) J. Mol. Biol. 328,
705–719.